This Week in Cardiology Podcast is Fixed
I did a deep dive into cardiac imaging, focusing on the use of CTA for the diagnosis of coronary artery disease. I think the anatomic imaging test is leading us astray.
This is a post about coronary artery imaging.
But first, in the matter of first world problems, the This Week in Cardiology podcast did not upload properly to the Apple and Spotify platforms yesterday. The Medscape team promptly fixed the issue and the podcast can now be heard in full.
I make note of this because I spent the week studying the use of coronary CT scanning.
Here are a few words about this relatively new way to assess for coronary artery disease. When you present for evaluation of chest pain, we try to diagnose and treat the problem. While chest pain can have many causes, a primary concern is coronary artery disease or CAD.
In days of old there was only “functional” testing for CAD. That is, we stressed the heart, either with exercise or vasodilators, and then assessed for ischemia (demand > supply due to coronary disease).
Inducible ischemia shows up on either the ECG, echo, or nuclear perfusion images. If positive, stress tests provided “indirect” evidence for coronary disease. The doctor could then treat the patient for CAD or refer for invasive coronary angiography—wherein contrast is injected into each of the three coronary arteries.
(Of course, the multimillion dollar ISCHEMIA trial found that there was no difference in outcomes for sending patients with positive stress tests immediately to the cath lab for coronary angiography vs initial medical treatment. But no matter, the train from positive stress tests to the cath lab remains at full capacity in the US.)
Coronary CT scans offer a different way to diagnose CAD. We call these “anatomic” tests—wherein the doctor can see the blockage as contrast goes down the coronary. (Note, a coronary CT is different from a coronary calcium scan. The latter simply measures the amount of calcium in the arteries whereas coronary CT is an actual angiogram of the inside of the vessel.)
The promise of coronary CT is that non-invasive angiography allows a more accurate diagnosis of CAD. Because you can see it.
I was part of a meta-analysis of studies comparing functional and anatomic testing in patients with chest pain.
We found that use of CTA decreased the incidence of MI but increased the use of coronary interventions. CTA vs functional stress testing did not result in a decrease in cardiac hospitalizations or mortality. However, and it was a big however, we were forced by reviewers to include a trial that was dissimilar from all the others.
The question we asked was which method was better to evaluate patients with chest pain—CTA or functional tests. We wanted to include only trials that directly compared the two testing strategies. Those trials, when combined, showed no difference in any outcome.
Peer reviewers, however insisted that we include the SCOT-HEART trial, which found that adding CTA to the normal routine of testing reduced MI in the future. Adding SCOT-HEART drove the positive results (MI) of our meta-analysis.
But there were two issues with SCOT-HEART. One was that it did not directly compare CTA and functional testing. Instead it was a comparison of CTA + functional testing to functional testing alone. So we did not originally include it in our meta-analysis. Peer reviewers insisted, and we had two choices: include it or have the paper rejected.
The second issue with SCOT-HEART arose when the authors presented the five year results, which showed dramatic reductions in MI when adding CTA.
The title of my commentary at that time was: Five Reasons I Don't Believe an Imaging Test Improves Outcomes. See also my post on SCOT-HEART on Sensible Medicine.
I discussed this issue again on this week’s podcast. The short story is that the degree of reduction in MI in SCOT-HEART was outsized and hard to explain.
I went further into CTA imaging on last week’s podcast.
For instance, the test is used in much different ways in the US vs Scotland. In cost-constrained health systems the test is used to decrease the number of expensive angiography and stent procedures. They do this by believing in the evidence and treating any CAD medically. Because that is what the evidence shows.
In the US, it’s the opposite. Hospitals are buying CTA scanners and putting them in ERs and chest pain centers, because the anatomic test finds CAD at higher rates than functional tests. It does not matter that patients can present with totally non-cardiac pain and the CTA finds coincidental CAD.
The culture here is that CAD should be investigated with an invasive angiogram and stented if significant—despite the evidence showing there is no benefit. I call the CTA scanners in the US cash machines because they send more patients to the cath lab than functional tests did.
The other issue I went into on the podcast was a new twist on using CTA called fractional flow reserve or FFR from the CT. This proprietary test from a company called HeartFlow purports to tell doctors whether a partial blockage seen on CTA is tight enough to limit flow. And if it is, that is bad, and it should be fixed.
Of course, FFR is used in the cath lab as well.
My take is that FFR is a flawed measure, no matter where it’s used. I went into detail on one of the seminal studies of CT-FFR, called the ADVANCE registry. I found a darn mess.
Conclusion
Many smart people have advised me not to be critical of a test. A test is just a test. The problem are the humans who misuse the test. That advice fits here. A CTA is a simple picture of coronary arteries. Our job is to use it wisely. That it is misused is the fault of humans not the scanner.
Coronary artery disease is part of the diverse disease of atherosclerosis. CAD treatment therefore should focus on treating atherosclerosis: diet, exercise, good control of blood pressure and cholesterol for instance.
Placing stents in stable partial blockages does not reduce the chance of future MI or death. It’s a hard concept but it’s been borne out by every clinical trial ever done.
If a CTA finds CAD the treatment is not a stent but medical treatment. Then if medical treatment fails, a stent can be considered.
Here are the various places that you can find the podcast. It publishes every Friday afternoon.
This Week in Cardiology podcast on Medscape (with a transcript)
On the one hand, I agree with the contention that the issue is not with the test per se, but with the end users of the results of that test (and the warped financial incentives that serve at cross purposes to proper outcome-driven patient care).
OTOH, it’s a failure of regulators and gate-keepers to authorize /permit/ give credence to a test that has not shown downstream endpoint benefits to begin with. It’s yet another example of regulators failing to properly discharge their fiduciary duties.
“Coronary artery disease is part of the diverse disease of atherosclerosis. CAD treatment therefore should focus on treating atherosclerosis: diet, exercise, good control of blood pressure and cholesterol for instance.
Placing stents in stable partial blockages does not reduce the chance of future MI or death. It’s a hard concept but it’s been borne out by every clinical trial ever done. “
I agree with all the above.
I would just add that it isn’t only the question of anatomically defining a blockage. It is beholden on us as physicians to assess the physiological cardiorespiratory and prognostic significance of that anatomical blockage in the anatomical context of its location. To give but one example, a high grade distal left main blockage might cause a balanced false negative nuclear perfusion scan, and there is not a shadow of a doubt that if it was a significant stenosis, and the patient was stressed using physiologic demand testing ( exercise dobutamine etc..), RATHER THAN flow related reactive vasodilator testing, that these would be profoundly positive. The reason we have wandered down this robotic road to cardiac catheterization is that we have used the wrong tests, and the third party payors funding these tests have been slavishly endorsing it. I refer to the all prevalent issue of vasodilator nuclear flow testing, using adenosine and similar vasodilator products as stressors. The compelling impetus for adopting this stress modality, is of course that these are revenue generating, well or at least adequately reimbursed, fast accomplished convenient office tests, which are money spinners for those who do them in the outpatient office, but generate a huge number of false positive tests, or even if they are true positive, often not really indicative of any real impetus to pursue cardiac catheterization, let alone an ‘on table’ need for immediate evascularization. Adding to this are vague definitions of what are ischemic symptoms. On any given Monday (at least where I practice), you will find 80% of the patients showing up in outpatient clinics for cardiac vasodilator stress tests do not have classical angina by the Levine criteria. They have some other variant frequently atypical variants of chest pain, and or dyspnea and risk factors which are then presented as the reasons for doing this test. However, we as clinicians have to ask. Are we doing the test to determine whether the pain is cardiac so that we can manage it that way, or are we doing the test to find an ischemic area and respond to it mechanically? You’re right to point out the ISCHEMIA trial and the great prognosis of patients, who can exercise to a good degree with or without developing ischemia. This is why we should be pushing wherever and whenever possible for physiologic testing, such as exercise or even dobutamine provocative ischemia testing. (I understand the limitations of dobutamine induced rate related double product, ischemia provocation). However, we blissfully ignore to a great degree the value and importance of the treadmill test, despite years of validated confirmed prognostic information for the Duke treadmill criteria. Third party payors should be encouraging (if not financially mandating) this for all patients who can exercise, and that too with an ischemia detection modality that doesn’t overemphasize the detection of small amounts of ischemia at the expense of triggering the increasingly unnecessary use of cardiac catheterizations, and subsequent ‘on table’ do or die interventions. Incidentally, an ‘on table’ intracoronary ultrasound or provocative FFR test may very well document some validated criteria for significant flow limitation of the stenosis. I don’t doubt this. The question really is even if it is flow limiting, IS the risk benefit ratio long-term for intervening on this stenosis at that time and location, worthwhile. There isn’t a whole host of information in this regard that goes out 10 to 20 years to guide us….
Ironically perhaps, one interesting use consideration of cardiac CT coronary angiography as a testing modality, is the intriguing idea that it could be used as a ‘next’ follow up test for a positive nuclear scan (unless the region of ischemia was suggestive of multivessel or left main disease). In this way, trivial or small to moderate amounts of ischemia (ISCHEMIA trial types) could be defined anatomically and managed medically. We could then avoid the use of on-table oculo-stenotic reflexes that
compel even the bravest of interventional cardiologists from walking away from a juicy high-grade lesion from fear or guilt of professional abdication. I speak unapologetically as a former interventional cardiologist. Just food for enormous thought 💭